Rishabh Rishabh (CDS’25,’28) (from left), Joseph Allen (CAS’27, Questrom’27), Rheona Mehta (CDS’26), Omar Augustin, a BU Spark! assistant solutions engineer, Aiganysh Ulanova (CAS’25, CDS’25), and Ann Liang (CDS’25) working on Team Weber’s presentation during Data Days for Good March 13.

Student Life

Data Days for Good Gets Students to Address Real-World Challenges

MassMutual and BU Spark! collaboration offers hands-on opportunity for Terriers to use their computer and data skills

Crunch time hit just after 1:30 last Thursday afternoon in Room 1148C in MassMutual’s offices high above Boston’s Seaport District.

The five BU students on Team Weber gathered in front of a big screen for a final run-through of their slide deck. They would have 10 minutes to present the results of four solid days of work.

Rheona Mehta (CDS’26) had recorded her video. Ann Liang (CDS’25) crunched the numbers one more time. Joseph Allen (CDS’27, Questrom’27) and Rishabh Rishabh (ENG’25,’28) tweaked their animated map. Aiganysh Ulanova (CAS’25, CDS’26) edited her introductory and closing remarks, getting their total time down to, well, 10 minutes-ish.

“Don’t worry, we’ve got this,” Ulanova told her teammates.

They were among five teams of BU students who spent their spring break participating in MassMutual’s Data Days for Good, applying their computer skills to cracking real-world problems. Team Weber’s assignment came from Boston City Councilor Ben Weber and concerned elderly housing in District 6. 

Bay State insurance giant MassMutual has been running the Data Days for Good program for its employees for years; this was the third year they sponsored an event for students in collaboration with the CDS Duan Family Spark! Initiative, the innovation and experiential learning lab housed at the Boston University Faculty of Computing & Data Sciences. Each team had mentors from MassMutual and BU Spark! to guide them as little or as much as they needed.

“We have this unique opportunity, through our partnership with MassMutual, to give students a concentrated, short, immersive experiential learning opportunity during spring break,” says Ziba Cranmer, founding director of BU Spark! “It’s a really innovative model of how to leverage a week and give students something they can put on their résumé with a known brand or reputable company and to actually do something that’s interesting in a really short period of time.”

“What we’re learning here,” Cranmer adds, “is how we apply our brilliant young minds full of computer skills to real-world problems that have a prosocial component.”

Starting on Monday, the five Team Weber students had to track down publicly available data on the elderly population and on housing in the district—both current and planned—and then merge all that into a single report. Not as easy as it sounds.

To do it, members of the team had to learn new skills in fuzzy matching, geocoding API, and shapefiles—and if you think we’re going to explain those techniques here, you’re reading the wrong story. What matters is that they worked.

“When we merged datasets, we started with 27 percent that were not matching” between datasets, using addresses and other data on latitude and longitude, Allen said during the team’s presentation. “With those techniques, we got that down to about 0.7 percent.”

“We are grateful that the students were able to take a look at housing data in District 6, so that our team can better understand the housing stock for our growing senior population,” says Bonnie Delaune, chief of staff for Weber. “Councilor Weber is focused on protecting and promoting the development of senior housing across the district.”

Other teams’ projects included one for the Hampden County Registry of Deeds, using machine learning to transcribe decades of old handwritten deeds and land ownership records into digital text, improving accessibility for residents of Hampden County. Another analyzed the demographics of nonprofit leadership for the New Commonwealth Fund, using LinkedIn data to identify trends in representation of people of color in executive and officer roles.

“Watching these presentations, I felt very proud of our students and what they can do when empowered with data and allowed to think outside the box,” says Azer Bestavros, associate provost for the Faculty of Computing & Data Sciences and a William Fairfield Warren Distinguished Professor. “It was inspiring to see their creativity at work and to see how resourceful they were. If this is what they can do in a few days over spring break working on Data Days for Good challenges, just imagine what they can do once they graduate!”

“Data Days for Good is a great example of what we can accomplish together through partnership across academia and industry,” says Alex Baldenko, head of data science at MassMutual. “It was inspiring to see how students and professionals worked together to bring ideas to life, and I know that we’re just scratching the surface of the impact we can have together.”

Different kinds of learning

“This experience in general is showing us what it’s like to work professionally with a client and even be in an office space like this,” Mehta says during a lunch break. “I think it’s showing us what our potential career path could look like, which is really valuable.”

The challenges are also different than in the classroom. “Projects we do back in our university are like a semester long. Here we have to do it in like a week’s time,” adds Rishabh.

“I’ve come to really love the intersection of public policy and technology just in general,” Mehta says. But “working with public data is like a skill set of its own, just because of how un-standardized it is.”

Their University mentor, Omar Augustin, a BU Spark! assistant solutions engineer, watched the team prepare, occasionally dishing out tips on everything from data filtering to the layout of a slide. 

Data scientist Audrey Bertin, their MassMutual mentor, says they didn’t need much help from her at all. “They know more than I do about a lot of the technology that they’re using,” Bertin says. “I just gave them life advice, about what things to look for when you’re getting into the data science space, what sort of skills to build in your portfolio.

“I feel like this project and event is honestly the perfect thing for that. So I was telling them, put this work on your public profile, and then you can talk about the real-life projects you’ve worked on.” That’s absolutely critical for getting jobs in data science, she says.

“They’re not fantastically adding to their technical skills,” Cranmer agrees. “What they are doing is learning what they can do with [those skills] and how to present those results. In class, they may get, you know, pretty sanitized, packaged problems that have answers. And the reality of doing data science is much, much messier. And so they’re in these experiential learning opportunities, which is what they get at BU Spark! broadly—they get thrown into the mess with no filters.

“And so that’s sort of the first thing. So they learn how to deal with reality—reality being things like messy datasets. And in order to do that, they have to learn how to merge one dataset to another and figure out what’s going to connect the two, so they can draw conclusions.”

And when a client in the real world wants results a certain way, “the students may have to learn new tools to be able to deliver,” Cranmer adds. “Because this field is constantly evolving, that’s actually a skill they need: ‘I don’t know how to do this. How do I find answers to how to do this? How do I learn that process and how do I apply it?’ And that happens often.”